CN111275806A - Parallelization real-time rendering system and method based on points - Google Patents
Parallelization real-time rendering system and method based on points Download PDFInfo
- Publication number
- CN111275806A CN111275806A CN201811379887.4A CN201811379887A CN111275806A CN 111275806 A CN111275806 A CN 111275806A CN 201811379887 A CN201811379887 A CN 201811379887A CN 111275806 A CN111275806 A CN 111275806A
- Authority
- CN
- China
- Prior art keywords
- point
- real
- point model
- module
- parallelization
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000009877 rendering Methods 0.000 title claims abstract description 88
- 238000000034 method Methods 0.000 title claims abstract description 49
- 230000001133 acceleration Effects 0.000 claims abstract description 29
- 238000005516 engineering process Methods 0.000 claims abstract description 21
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 17
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 11
- 238000009826 distribution Methods 0.000 claims abstract description 4
- 238000004364 calculation method Methods 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 11
- 238000005286 illumination Methods 0.000 claims description 8
- 238000005070 sampling Methods 0.000 claims description 5
- 238000010425 computer drawing Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 description 6
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000013139 quantization Methods 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006837 decompression Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 239000012265 solid product Substances 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/10—Geometric effects
- G06T15/20—Perspective computation
- G06T15/205—Image-based rendering
Abstract
The invention discloses a parallelization real-time rendering system based on points, wherein a real-time efficient point model rendering module is electrically connected with a parallelization and distribution type point model rendering module, the parallelization and distribution type point model rendering module is electrically connected with a real-time point model rendering module based on hardware acceleration, the real-time point model rendering module based on hardware acceleration is electrically connected with a task decomposition module drawn by a point model, the task decomposition module drawn by the point model is electrically connected with a parallelization k-nearest search module, and the parallelization k-nearest search module is electrically connected with a ray tracking algorithm module based on GPU hardware acceleration; the method realizes the real-time and efficient point model rendering based on the parallelization and distributed point model rendering, the real-time point model rendering based on hardware acceleration, the task decomposition technology of the point model drawing, the parallelization k-nearest search technology and the ray tracing algorithm based on GPU hardware acceleration.
Description
Technical Field
The invention belongs to the field of rendering systems, and particularly relates to a parallel real-time rendering system based on points. Meanwhile, the invention also relates to a method for parallelizing the real-time rendering system based on the points.
Background
In the field of real-time rendering of computer graphics, model rendering using a triangular mesh as a primitive is the most popular method in current interactive graphics rendering due to the advantages of simplicity, strong expressive force, good hardware support and the like, and is also the most mature method at present. With the rapid development of computer graphics theory and technology, the application field of computer graphics has spread to many aspects such as scientific exploration, engineering design, machine manufacturing, simulation, medicine and health, cultural relic protection, game entertainment and the like, and meanwhile, one of the characteristics is that three-dimensional geometric data is widely used to describe scenes, and the scale and complexity of the represented scenes are also increased sharply. In recent years, with the refinement of scene modeling, we increasingly use a large-scale, highly complex polygon model, which causes the overhead for managing the connection relationship of the polygon model to increase greatly. A model may contain up to millions of triangles, and the projected area of each triangle in screen space may occupy only a few pixels, even within one pixel. In such a case, the rasterization operation is no longer as efficient.
In the aspect of acquiring real three-dimensional information, with the increasing popularization and application of 3D digital scanning technology and scanners in recent years, users can acquire a large amount of three-dimensional and image data related to objects in a short time. The 3D scanner outputs a large number of closely spaced sample points of the target object, which have no topological relationship to each other. In the traditional method, discrete sampling points obtained by scanning need to be converted into a triangular model for further processing and drawing, but with the continuous expansion of the scale of sampling data, the method for converting or reconstructing the polygonal model becomes more and more inefficient, and the limitation of the triangular representation method becomes more obvious.
Therefore, new fields and new technologies are emerging that present significant challenges to conventional graphics processing rendering techniques and methods.
Among the point-based rendering methods, the foreign technologies mainly include the qssplat method and the Splatting method.
The qssplat method is proposed in the article published by legoy et al at the university of standard on SIGGRAPH in 2000, and uses a tree-like hierarchical bounding sphere data structure to complete tasks such as visibility selection, LOD control, real-time drawing and the like, wherein each node in the tree contains information related to a sphere, and the size of each node is 48 bits. And traversing the hierarchical tree according to a breadth-first method during rendering. For each intermediate node, it is first determined whether the ball is completely off-screen or completely back-facing. If so, the node is not visible, the node and all of its nodes are ignored, and this enables visibility selection. If at least a part of the node is visible, comparing the projection size of the node on the screen with a threshold value, and if the projection size of the node is larger than the threshold value, continuing to recurse downwards; if it is less than the threshold or has reached a leaf node, then a "Splat" is drawn at the location and size on the screen determined by the sphere location and radius of the node. The size of the threshold value can be dynamically determined by the drawing time of the last frame of image so as to meet the requirement of drawing the number of frames per second specified by a user. Once the user stops moving the mouse, a smaller threshold value is gradually used for drawing until the size of the leaf node or Splat is one pixel point.
The highly compressed data structure is a big feature of the qssplat method, in which Levoy, etc. quantizes each parameter information to be stored, and selects the quantization order of magnitude that balances the drawing speed and the image quality through experimental comparison. Through the quantization processing, the number of bytes occupied by the information stored in each node is greatly compressed, and meanwhile, the error is controlled within an acceptable range. The representation modes of the compressed Qsplat data in the memory and the external memory are the same, so that the work of decompression is not needed to be carried out firstly during drawing, and only the data is required to be directly called into the memory, thereby further accelerating the drawing speed.
Since qssplat is mainly directed to the fast display of scanned large data volume models, the required data file for each model is still large despite the considerable compression rate, and especially when data is remotely called in through a network, it is likely that the data transmission cannot keep pace with the display speed. To solve this problem, the streamingqssplat method modifies the qssplat method. Because qssplat can be terminated at any level in the recursion process, the modification of the Streaming qssplat method employs a very simple, straightforward approach: once the desired data has not been passed, the recursion is aborted. Therefore, a coarser image can be displayed in advance, and the data are refined step by step after being transmitted. The Streaming Qsplat method adopts a GPU hardware acceleration technology, and improves the drawing speed by about two orders of magnitude. The reported rate is plotted at about 50M points per second.
The Splatting technology is a mainstream technology of current point drawing, Surfel is taken as a drawing element of a point model, each point is taken as a thin wafer, Surfel of each point is mutually overlapped to form a compact object surface, and a smooth image is obtained through visibility preprocessing image reconstruction. The radius of the disk is related to the curvature of the curved surface at that point. The drawing method is to project the whole Surfel onto a screen so as to realize visibility judgment and image reconstruction at the same time. The core problems include: and determining the projection of the Surfel on the screen and determining the mutual occlusion relation and the mutual fusion relation between the projections. Zwicker introduces a signal processing method into point drawing, and adopts an EWA anti-aliasing filtering technology to obtain a high-quality drawing result. Ren and Mario deduces and simplifies the EWA filter operator of the object space on the basis, and obtains higher drawing efficiency by utilizing the latest graphics hardware acceleration technology. The reported plotting rate is 1.5M points per second.
In both these rendering techniques, qsplit focuses on rendering efficiency, but has the disadvantage of severe aliasing. The concept of signal processing is applied in Splatting, resampling and filtering in screen space are ingeniously combined together, the aliasing problem of point drawing is well solved, a high-quality drawing effect is obtained, and the method has the following defects: multiple renderings are required, the translucent volume cannot be rendered realistically, and the like. The prior art is more multilaterally focused on point model rendering with a single machine. During single-point model rendering, partial optimization algorithm is realized by using GPU acceleration.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides a parallel real-time rendering system based on points and a method thereof.
In order to achieve the purpose, the invention adopts the following technical scheme:
the utility model provides a parallelization real-time rendering system based on point, includes real-time efficient point model rendering module, real-time efficient point model rendering module electric connection has parallelization, the point model rendering module of distributing type, parallelization, the point model rendering module electric connection of distributing type has real-time point model rendering module based on hardware acceleration, real-time point model rendering module electric connection has the task decomposition module that the point model was drawn based on hardware acceleration, the task decomposition module electric connection that the point model was drawn has parallelization k-nearest search module, parallelization k-nearest search module electric connection has ray tracing algorithm module based on GPU hardware acceleration.
Preferably, the real-time efficient point model rendering module adopts direct drawing of point data to omit the original surface reconstruction process, and uses the information of the point for direct drawing.
Preferably, the point model rendering technology adopted by the parallelization and distributed point model rendering module divides a total task into a plurality of tasks by using a certain task parallelization strategy, reasonably distributes the tasks to each calculation node, and synthesizes a final result after each task is completed.
Preferably, the real-time point model rendering module based on hardware acceleration introduces GPU calculation, fully utilizes the powerful operation function of the GPU, and solves the problem of point model drawing.
Preferably, in the parallelization of the point model, the task decomposition module for drawing the point model must decompose the task into a plurality of small subtasks, and each computing node calculates the small subtasks and then combines the plurality of subtasks into a final result.
Preferably, in the computer drawing of the point model by the parallelized k-nearest search module, the sense of reality comes from the calculation of illumination, and one of the most important input information of the illumination calculation is a normal vector of a point, which is essential information in the surface processing process and is estimated by adopting a neighborhood of a sampling point.
Preferably, the ray tracing algorithm module based on the GPU hardware acceleration adopts the GPU to take charge of main calculation, and the CPU takes charge of control and coordination of calculation.
The invention also provides a method for parallelizing real-time rendering system based on points, which comprises the following steps:
s1, directly drawing point data through a real-time efficient point model rendering module;
s2, when the model point data is projected to a screen space to search k-nearest points, after the k-nearest points are searched, calculating an average value through texture output, calculating the color of a normal vertex, and then drawing Splat in a frame buffer through the texture;
s3, when the model point data is projected to a point in the screen space within a predetermined range, Splat is drawn in the frame buffer by texture output.
Compared with the traditional solid product, the parallelization real-time rendering system based on the points and the method thereof provided by the invention have the following advantages:
1. according to the method, parallelization and distributed point model rendering, real-time point model rendering based on hardware acceleration, a task decomposition technology of point model drawing, a parallelization k-nearest search technology and a ray tracing algorithm based on GPU hardware acceleration are adopted to realize real-time and efficient point model rendering;
2. rendering a point model in real time and efficiently: due to the fact that points are not topological, triangle reconstruction has a lot of uncertainty, the original surface reconstruction process is omitted through direct drawing of the point data, the point information is used for direct drawing, the period of virtual reality modeling is simplified, and workload is reduced;
3. parallelized, distributed point model rendering: because the point cloud scale is huge, a single computer is difficult to effectively process a large-scale point model, the problem can be effectively solved by adopting a parallelization distributed processing platform, and the drawing based on a parallel architecture is adopted for the traditional single-machine drawing of the point model;
4. real-time point model rendering based on hardware acceleration: the strong operation function of the GPU is fully utilized, the drawing problem of the point model is solved, the method is different from the traditional single calculation mode of the CPU, and the introduction of the GPU calculation greatly improves the rendering efficiency and performance of the point model.
Drawings
FIG. 1 is a block diagram of a specific computing architecture of the present invention;
FIG. 2 is a block diagram of a task scheduling system according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to fig. 1 and fig. 2 in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. The specific embodiments described herein are merely illustrative of the invention and do not delimit the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The utility model provides a parallelization real-time rendering system based on point, includes real-time efficient point model rendering module, real-time efficient point model rendering module electric connection has parallelization, the point model rendering module of distributing type, parallelization, the point model rendering module electric connection of distributing type has real-time point model rendering module based on hardware acceleration, real-time point model rendering module electric connection has the task decomposition module that the point model was drawn based on hardware acceleration, the task decomposition module electric connection that the point model was drawn has parallelization k-nearest search module, parallelization k-nearest search module electric connection has ray tracing algorithm module based on GPU hardware acceleration.
Furthermore, the real-time and efficient point model rendering module adopts direct drawing of point data to omit the original surface reconstruction process, and utilizes the information of the points to draw directly.
By adopting the technical scheme: the basic modeling element of the virtual reality method based on the triangle (polygon) is the triangle, any object must be composed of triangles with different sizes, along with the rapid development of the three-dimensional measurement technology, particularly the laser three-dimensional measurement technology, great progress is made in the aspect of obtaining the three-dimensional information of the object, a large amount of point clouds with high precision and density can be obtained in a very short time, if the object is displayed by utilizing the scanning number, the data must be converted into polygon data, because the points have no topology, the triangle reconstruction is provided with much uncertainty, the original surface reconstruction process is omitted by directly drawing the point data, the information of the points is directly drawn, the period of virtual reality modeling is simplified, and the workload is reduced.
Furthermore, the point model rendering technology adopted by the parallelization and distribution type point model rendering module divides a total task into a plurality of tasks by using a certain task parallelization strategy, reasonably distributes the tasks to each calculation node, and synthesizes a final result after each task is completed.
By adopting the technical scheme: because of the huge point cloud scale, a single computer is difficult to effectively process a large-scale point model, the problem can be effectively solved by adopting a parallelization distributed processing platform, the adopted point model drawing technology divides a total task into a plurality of tasks by utilizing a certain task parallelization strategy, the tasks are reasonably distributed to each calculation node, and a final result is synthesized after each task is completed.
Furthermore, the real-time point model rendering module based on hardware acceleration introduces GPU calculation, fully utilizes the powerful operation function of the GPU, and solves the problem of point model drawing.
By adopting the technical scheme: in the current popular GPU, a vertex shader and a pixel shader which are both programmable controllers are provided, and are a typical stream processor, the stream processor is mainly different from the traditional vector processor in that the stream processor does not have a large-capacity memory and can read and write, only a temporary register is directly used for performing stream data operation on a chip, and due to the ultra-long pipeline and the parallel computing function of the GPU, the operation speed of the GPU is very high, and the project fully utilizes the powerful operation function of the GPU to solve the drawing problem of a point model.
Further, in the parallelization of the point model, the task decomposition module for drawing the point model must decompose the task into a plurality of small subtasks, and each computing node calculates the small subtasks and then combines the plurality of subtasks into a final result.
Adopt above-mentioned technical scheme: in the parallelization of the point model, a task must be decomposed into a plurality of small subtasks, each computing node computes the small subtasks and then combines the small subtasks into a final result, the key of the point model is the task decomposition, the key of the task decomposition is to find a suitable point model representation data structure, the parallelization operation can be effectively carried out, and besides the parallelization is easy, the data structure adopted in the task decomposition must have higher storage efficiency and higher compression ratio relative to source data, so that the communication between the computing nodes is facilitated.
Furthermore, in the computer drawing of the point model by the parallelized k-nearest search module, the sense of reality comes from the calculation of illumination, and one of the most important input information of the illumination calculation is a normal vector of a point, which is essential information in the surface processing process and is estimated by adopting the neighborhood of the sampling point.
By adopting the technical scheme: in the computer drawing of a point model, the sense of reality comes from the calculation of illumination, and one of the most main input information of the illumination calculation is a normal vector of a point, which is essential information in the surface processing process.
Further, the ray tracing algorithm module based on the GPU hardware acceleration adopts the GPU to take charge of main calculation, and the CPU is responsible for control and coordination of calculation.
Through the technical scheme: the GPU is employed to take care of the main computations, while the CPU is responsible for the control and coordination of the computations. Because the data used in the GPU is texture, there are problems how to express and generate rays in the GPU, how to organize point model structures in the GPU, and how to perform intersection operations between rays and point models.
The invention also provides a method for parallelizing real-time rendering system based on points, which comprises the following steps as shown in figure 1:
s1, directly drawing point data through a real-time efficient point model rendering module;
s2, when the model point data is projected to a screen space to search k-nearest points, after the k-nearest points are searched, calculating an average value through texture output, calculating the color of a normal vertex, and then drawing Splat in a frame buffer through the texture;
s3, when the model point data is projected to a point in the screen space within a predetermined range, Splat is drawn in the frame buffer by texture output.
In summary, the following steps: compared with the traditional solid-state product, the parallelization real-time rendering system and the parallelization real-time rendering method based on the points have the advantages that the real-time efficient point model rendering is realized: due to the fact that points are not topological, triangle reconstruction has a lot of uncertainty, the original surface reconstruction process is omitted through direct drawing of the point data, the point information is used for direct drawing, the period of virtual reality modeling is simplified, and workload is reduced;
parallelized, distributed point model rendering: because the point cloud scale is huge, a single computer is difficult to effectively process a large-scale point model, the problem can be effectively solved by adopting a parallelization distributed processing platform, and the drawing based on a parallel architecture is adopted for the traditional single-machine drawing of the point model;
real-time point model rendering based on hardware acceleration: the strong operation function of the GPU is fully utilized, the drawing problem of the point model is solved, the method is different from the traditional single calculation mode of the CPU, and the introduction of the GPU calculation greatly improves the rendering efficiency and performance of the point model.
Therefore, the invention realizes the real-time and high-efficiency point model rendering by parallelizing and distributing point model rendering, real-time point model rendering based on hardware acceleration, task decomposition technology of point model drawing, parallelizing k-nearest search technology and ray tracing algorithm based on GPU hardware acceleration.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (8)
1. A parallelized real-time rendering system based on points is characterized in that: the system comprises a real-time efficient point model rendering module, wherein the real-time efficient point model rendering module is electrically connected with a parallelization and distributed point model rendering module, the parallelization and distributed point model rendering module is electrically connected with a hardware acceleration-based real-time point model rendering module, the hardware acceleration-based real-time point model rendering module is electrically connected with a task decomposition module drawn by a point model, the task decomposition module drawn by the point model is electrically connected with a parallelization k-nearest search module, and the parallelization k-nearest search module is electrically connected with a GPU hardware acceleration-based ray tracking algorithm module.
2. The point-based parallelized real-time rendering system of claim 1, wherein: the real-time efficient point model rendering module adopts direct drawing of point data to omit the original surface reconstruction process, and utilizes the information of the points to draw directly.
3. The point-based parallelized real-time rendering system of claim 1, wherein: the point model rendering technology adopted by the parallelization and distribution type point model rendering module divides a total task into a plurality of tasks by utilizing a certain task parallelization strategy, reasonably distributes the tasks to each calculation node, and synthesizes a final result after each task is completed.
4. The point-based parallelized real-time rendering system of claim 1, wherein: the real-time point model rendering module based on hardware acceleration introduces GPU calculation, makes full use of the powerful operation function of the GPU, and solves the problem of point model drawing.
5. The point-based parallelized real-time rendering system of claim 1, wherein: in the parallelization of the point model, the task decomposition module for drawing the point model must decompose the task into a plurality of small subtasks, and each computing node calculates the small subtasks and then combines the plurality of subtasks into a final result.
6. The point-based parallelized real-time rendering system of claim 1, wherein: in the computer drawing of the point model by the parallelized k-nearest search module, the sense of reality comes from the calculation of illumination, and one of the most main input information of the illumination calculation is a normal vector of a point, which is essential information in the surface processing process and is estimated by adopting the neighborhood of a sampling point.
7. The point-based parallelized real-time rendering system of claim 1, wherein: the ray tracing algorithm module based on GPU hardware acceleration adopts a GPU to take charge of main calculation, and a CPU is used for controlling and coordinating calculation.
8. A method of point-based parallelized real-time rendering system according to claim 1, characterized in that: the method comprises the following steps:
s1, directly drawing point data through a real-time efficient point model rendering module;
s2, when the model point data is projected to a screen space to search k-nearest points, after the k-nearest points are searched, calculating an average value through texture output, calculating the color of a normal vertex, and then drawing Splat in a frame buffer through the texture;
s3, when the model point data is projected to a point in the screen space within a predetermined range, Splat is drawn in the frame buffer by texture output.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811379887.4A CN111275806A (en) | 2018-11-20 | 2018-11-20 | Parallelization real-time rendering system and method based on points |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811379887.4A CN111275806A (en) | 2018-11-20 | 2018-11-20 | Parallelization real-time rendering system and method based on points |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111275806A true CN111275806A (en) | 2020-06-12 |
Family
ID=71002879
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811379887.4A Pending CN111275806A (en) | 2018-11-20 | 2018-11-20 | Parallelization real-time rendering system and method based on points |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111275806A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114827032A (en) * | 2021-01-20 | 2022-07-29 | 辉达公司 | Performing network congestion control with reinforcement learning |
| CN117830489A (en) * | 2024-03-05 | 2024-04-05 | 浙江小牛哥科技有限公司 | Intelligent indoor design image rendering system |
| CN117830489B (en) * | 2024-03-05 | 2024-05-03 | 浙江小牛哥科技有限公司 | Intelligent indoor design image rendering system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5821950A (en) * | 1996-04-18 | 1998-10-13 | Hewlett-Packard Company | Computer graphics system utilizing parallel processing for enhanced performance |
| CN103049927A (en) * | 2013-01-17 | 2013-04-17 | 浙江大学 | Real-time ray tracing and rendering method based on GPU (Graphics Processing Unit) aggregate |
| US20140198097A1 (en) * | 2013-01-16 | 2014-07-17 | Microsoft Corporation | Continuous and dynamic level of detail for efficient point cloud object rendering |
| CN107506306A (en) * | 2017-09-28 | 2017-12-22 | 网易（杭州）网络有限公司 | The method of testing and device of fine arts resource |
| CN108765539A (en) * | 2018-05-24 | 2018-11-06 | 武汉斗鱼网络科技有限公司 | Image rendering method, device, equipment and storage medium based on OpenG L ES |
-
2018
- 2018-11-20 CN CN201811379887.4A patent/CN111275806A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5821950A (en) * | 1996-04-18 | 1998-10-13 | Hewlett-Packard Company | Computer graphics system utilizing parallel processing for enhanced performance |
| US20140198097A1 (en) * | 2013-01-16 | 2014-07-17 | Microsoft Corporation | Continuous and dynamic level of detail for efficient point cloud object rendering |
| CN103049927A (en) * | 2013-01-17 | 2013-04-17 | 浙江大学 | Real-time ray tracing and rendering method based on GPU (Graphics Processing Unit) aggregate |
| CN107506306A (en) * | 2017-09-28 | 2017-12-22 | 网易（杭州）网络有限公司 | The method of testing and device of fine arts resource |
| CN108765539A (en) * | 2018-05-24 | 2018-11-06 | 武汉斗鱼网络科技有限公司 | Image rendering method, device, equipment and storage medium based on OpenG L ES |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114827032A (en) * | 2021-01-20 | 2022-07-29 | 辉达公司 | Performing network congestion control with reinforcement learning |
| CN117830489A (en) * | 2024-03-05 | 2024-04-05 | 浙江小牛哥科技有限公司 | Intelligent indoor design image rendering system |
| CN117830489B (en) * | 2024-03-05 | 2024-05-03 | 浙江小牛哥科技有限公司 | Intelligent indoor design image rendering system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8570322B2 (en) | Method, system, and computer program product for efficient ray tracing of micropolygon geometry | |
| Borgeat et al. | GoLD: interactive display of huge colored and textured models | |
| US8725466B2 (en) | System and method for hybrid solid and surface modeling for computer-aided design environments | |
| US10997774B2 (en) | Layered, object space, programmable and asynchronous surface property generation system | |
| US20100091018A1 (en) | Rendering Detailed Animated Three Dimensional Characters with Coarse Mesh Instancing and Determining Tesselation Levels for Varying Character Crowd Density | |
| US9684997B2 (en) | Efficient rendering of volumetric elements | |
| US11816783B2 (en) | Enhanced techniques for traversing ray tracing acceleration structures | |
| US11373358B2 (en) | Ray tracing hardware acceleration for supporting motion blur and moving/deforming geometry | |
| JP5055214B2 (en) | Image processing apparatus and image processing method | |
| Wand et al. | Interactive Editing of Large Point Clouds. | |
| CN101119485A (en) | Characteristic reservation based three-dimensional model progressive transmission method | |
| CN113034656B (en) | Rendering method, device and equipment for illumination information in game scene | |
| US11508112B2 (en) | Early release of resources in ray tracing hardware | |
| KR20080018404A (en) | Computer readable recording medium having background making program for making game | |
| Zach et al. | Time-critical rendering of discrete and continuous levels of detail | |
| Ma et al. | An integrated VR–GIS navigation platform for city/region simulation | |
| CN115131482A (en) | Rendering method, device and equipment for illumination information in game scene | |
| Xue et al. | Efficient GPU out-of-core visualization of large-scale CAD models with voxel representations | |
| CN111275806A (en) | Parallelization real-time rendering system and method based on points | |
| Bao et al. | Billboards for tree simplification and real-time forest rendering | |
| US11948338B1 (en) | 3D volumetric content encoding using 2D videos and simplified 3D meshes | |
| Schmalstieg | A survey of advanced interactive 3-d graphics techniques | |
| Ruiz et al. | Exploitation of Interframe Redundancy for Real-Time Volumetric Reconstruction of Arbitrary Shapes | |
| Blais et al. | GoLD: Interactive Display of Huge Colored and Textured Models | |
| Domaratius | Chemnitz, January 25, 2007 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200612 |